Activity Number:
|
79
|
Type:
|
Contributed
|
Date/Time:
|
Monday, August 12, 2002 : 8:30 AM to 10:20 AM
|
Sponsor:
|
Biometrics Section*
|
Abstract - #301236 |
Title:
|
Estimating the Difference in the Mean Medical Costs Between Two Groups in Observational Studies with Right-censored Data
|
Author(s):
|
Pei-Yun Chen*+ and Anastasios Tsiatis
|
Affiliation(s):
|
Merck Research Laboratories and North Carolina State University
|
Address:
|
P.O. Box 2000, RY34-A316, Rahway, New Jersey, 07065,
|
Keywords:
|
Causal inference ; Counting process ; Cox's proportional hazards model ; Cost analysis ; Observational study ; Survival analysis
|
Abstract:
|
Comparison of medical costs between two groups is often of interest in biomedical studies, because proper cost analysis may lead to more cost-efficient medical interventions. But it is often complicated because of the fact that the response of interest may be right-censored due to incomplete follow-up. Additionally, in observational studies where patients receiving one treatment are not guaranteed to be prognostically comparable to those receiving the other, we also need to account for treatment imbalances in confounding factors. In this article, we propose an estimator and its asymptotic variance for the difference in the mean medical costs between two groups that accounts for treatment imbalances. Simulation studies are conducted to show that our estimator performs well in finite samples. We apply our method to an observational database of acute coronary syndrome patients from Duke University Medical Center to estimate the treatment effect on the medical costs over five years.
|